AIRBNB RENTAL PRICES ANALYSIS- Montreal Vs Toronto

Introduction

This project aims at evaluating data from Montreal and Toronto AirBnBs, that was retrieved from InsideAirBnB in November 2020. Inside AirBnB gathers and collects publicly available data about a city's AirBnB listings. Among the data provided is a detailed overview of property listings, calendar availability and AirBnB stay reviews.

I followed the CRISP-DM process for this project.

Table of Contents

I- Business Understanding

I will handle the following questions:

II- Data Understanding and Preparation

First, let's read in the necessary libraries and datasets.

1. Listings File

Notes

Notes

In both Montreal and Toronto, "Entire home/apt" is the most common room type. Other room types are "Private room", "Shared room", and "Hotel room".

2.Reviews File

Notes

Reviews dataset lists individual reviews for the listings. Each row includes a listing id, review id, date, reviewer id, reviewer name, and the review text.

3. Calendar File

Notes

This dataset show the availability of listings on certain dates. Availability is encoded as f or t in the "available" column.

The calendar provides information on the price for staying at this AirBnB, and the number of minimum and maximum nights per stay.

To sum up:

Cleaning data

There are just some issues that needs to be fixed

Drop unnecessary columns

Handling erroneous datatypes :

Convert any column date from string to dateformat

Handling missing values

Additional cleaning

III- Exploratory Data Analysis

1.How long have hosts been listing properties on Airbnb in both Montreal and Toronto?

Notes

The oldest Montreal and toronto listing was first listed on the site in August 2008. From 2011 onwards, the number of listings started increasing considerably. However, growth in the number of new hosts (of those currently listing on the site) has been decreasing since the second half of the year 2016.The significant decrease was observed in 2020 due to the covid 19 pandemic.

2. How many days a year do homeowners make their homes available to rent - Montreal Vs Toronto?

Notes

The way people use their homes for rentals in Montreal and Toronto is pretty similar, where a majority of people hardly ever rent out their homes at all.

3. How much do people charge to rent their homes? How does this compare from Montreal to Toronto?

Notes

On average prices in Toronto are just more expensive than in Montreal all year round.

In both cities, prices spike on weekends. The price curves are relatively flat, except the price increase during Christmas season. In Toronto, prices spike more than 40 dollars on average from the beginning to end of December, in Montreal around 30 dollars. Another exception is the upward tendency during the spring and summer months.

Notes

From the above, , you can see that the distribution of the average listing prices appears to be pretty similar between Montreal and Toronto homes. However, the Toronto homes are more expensive on average (higher maximum and average mean and median prices).

4.What is the ideal time to visit Montreal and Toronto ?

Notes

In both Montreal and Toronto, the proportion of available listings for any given date never rises above 50%. The calendar data was retrieved in November 2020, so it is no surprise, that the curve starts close to fully not available. Until December, the curves rise to almost 50% in both Toronto and Montreal. In January 2021, the availability decreases again in the form of stairs. The first huge decrease is at the beginning of January but remain almost flat until spring before another jump to over 70% occurs in both cities.

Of course, that holiday season (Christmas time) and the summer are the best time to visit both cities. Nevertheless, if you are searching for lower prices and more choices of AirBnBs, autumn and winter (except December) are the ideal time to visit these beautiful cities.

Let's categorise prices Let price less than 75 dollars be low, between 75 and 300 dollars be medium and above 300 dollars be high

5.Which Neighborhood is the most rated? Popularity of airbnb by location

Notes

In Toronto, "Waterfront Communities-The Island" is the neighborhood that has by far the highest number of Airbnb listings. In Montreal, the highest AirBnB density is in Ville-Marie, followed by Le Plateau Mont Royal. Staten Island has the least amount of Airbnbs.

Notes

Waterfront, Annex, Niagara, Bay Street Corridorand Church yonge have the higher average location scores than the highest rated Montreal neighborhood (Le plateau Mont Royal). In Toronto, "Waterfront Communities-The Island" receives the highest score because it closes to the most famous city hotspots (CN Tower, Toronto International Film Festival, Toronto Island Park and beaches on the park…) and has the more convenient transportation (subway streetcars, ferries). While in Montreal, “Le plateau Mont-Royal » receives the highest score, it’s the most densely populated borough in Canada.

6.Which areas of Montreal and Toronto are the most expensive and which area is the best?

Notes

It is obvious that the highly rated locations would also tend to be costly this is because in the case of constant supply, the higher the demand, the higher the price is.

In Toronto, the 5 evaluated neighborhoods (top 5 with the highest number of Airbnb listings) are on the more expensive side, especially in the “Waterfront communities” averaging at 194.37 dollars. For low budget travelers, there are some chances to find low-priced listings in “Annex” and “Church-Yonge Corridor”.

In Montreal, “Ville marie” and “Le plateau Mont royal” have the most balanced distribution of lower price and medium to high price listings. On average, “ville marie” seems like relatively the priciest of all (the top 5 analysed), averaging at 134 dollars. For low budget travelers, there are chances to find low-priced listings in “Rosemont La Petite Patrie”, “Cote des neiges notre Dame de Grace” and “le Sud Ouest”.

7Most common room types in Montreal and Toronto

Now that we have an idea of the availability, prices of homes and best neighbors, I would look at other characteristics of listings.

Notes

Almost listings, in both Montreal and Toronto are entire homes (renting the entire property). Most of the remaining ones are private rooms (renting a bedroom and possibly also a bathroom, but there will be other people in the property). Fewer than 2% are shared rooms (sharing a room with either the property owner or other guests).

8. What are the most common words used to describe a listing? Are the same words used for Montreal and Toronto homes?

I tried to find words that describe a stay that were used across both Montreal and Toronto. Overall it seems that both locations sound like great places to stay at an AirBnB.

Montreal "apartment"s are "clean" "place"s to "stay". Guests appreciate "everything", "confortable room"s, "great location"s and "host"s. They also mention the "subway".

Toronto "apartment"s are "clean" "place"s to "stay". Here, "great location", "great place and the "helpful amd great host" are mentioned in this sequence. Again, they would " highly recommend" the "apartment", where "everything" has been "comfortable".

IV- Data modelling and Evaluation

The case of Montreal

Notes

There are some colinearity between certain variables, as an example, beds, bedrooms and accomodates. All collinear variables would be dropped in the next step.

Linear regression

Notes

The residuals histogram looks normally distributed.

Ridge Model
Lasso model
Decision Tree Regression
Random Forest regression

Notes

We can see that Random Forest model has the highest R^2 value(92%) and least Mean Squared Error. Hence it is the best model.

What are the main factors that influence Airbnb renting price?

Notes

The most important feature is how many people the property accommodates, it's the first thing you would use to search for properties with in the first place. It is also not surprising that features related to location (latitude and longitude) and score/reviews are in the top ten. Both higher ratings and more numerous reviews can lead to a higher price. The “Age/host duration” variable, namely, the duration of time since the listing was established, is the fourth most important feature affecting Airbnb listing prices.

The case of Toronto

Notes

We can see that Random Forest model has the least error. Hence it is the best model.

What are the main factors that influence Airbnb renting price?

Conclusion

In this article, we took a look at AirBnB data from Montreal and Toronto to understand some areas of interest, especially pricing, following the Cross Industry Standard Process for Data Mining (CRISP-DM) process.

There were some useful insights found at each level, that helped us to identify some factors that can impact an Airbnb listing’s price, by using five regression models.

This work could be improved by implementing a recommendation system which suggests an airbnb listing using the keywords the user has provided. It would return the most relevant airbnb listing.